all repos — openbox @ 1b13d34c1e0ef69036de51c289a83da3aaecef34

openbox fork - make it a bit more like ryudo

i think im handling reparent events correctly now, by unmanaging the window and then XMapping the client
Dana Jansens danakj@orodu.net
commit

1b13d34c1e0ef69036de51c289a83da3aaecef34

parent

e5eb783aa22440625d497a07fa079cae93bdaa86

2 files changed, 12 insertions(+), 2 deletions(-)

jump to
M src/client.ccsrc/client.cc

@@ -1269,6 +1269,12 @@ the window manager never sees it because an unmap event is not sent

to an already unmapped window. */ + // put another copy of this event on the stack (for the frame reparent + // process) + XEvent ev; + ev.xreparent = e; + XPutBackEvent(otk::OBDisplay::display, &ev); + // this deletes us etc Openbox::instance->screen(_screen)->unmanageWindow(this); }
M src/frame.ccsrc/frame.cc

@@ -436,8 +436,12 @@ // check if the app has already reparented its window to the root window

XEvent ev; if (XCheckTypedWindowEvent(otk::OBDisplay::display, _client->window(), ReparentNotify, &ev)) { - // XXX: ob2/bb didn't do this.. look up this process in other wm's! - //XPutBackEvent(otk::OBDisplay::display, &ev); + /* + If the app reparented itself, then we unmanage the window. This causes + the window to be unmapped, so to be nice to it, we remap the window + here. We don't put the event back onto the stack because we put it there + in the first place. + */ XMapWindow(otk::OBDisplay::display, _client->window()); } else { // according to the ICCCM - if the client doesn't reparent to